is 3% swr the new 4% ,an interview with dr wade pfau (grandparent, grandpa)

[mathjak107]

you are not following .

typically retirement planning standardizes on planning to 95 .. if you are first retiring and 80 years old your planning time frame would be about 15 years.

to have your money last 15-20 years is very different than the planning needed to last 30-40 years.

15 YEARS IS A LOT LESS DEPENDENANT ON SEQUENCES,INFLATION AND YEARLY RETURNS .

the longer you have to plan for the more critical those 3 issues become.
you still need to watch what you spend for shorter time frames but the 3 parameters are less critical to your outcome.

[mathjak107]

a picture helps illustrate :

if all you need is a 15 year time frame to get to age 95 a fairly safe withdrawal rate has been 6% using only bonds . it has failed only 11% of the time. 5% withdrawal rate over 15 years never failed with just bonds.

but using bonds only out to 30 years at 6% withdrawal rates has failed 96% of every time frame.

taking a 3% withdrawal rate bonds met every time frame out to 15 years just great.

but going out 30 years it failed 20% of the time frames and 31% of the time out to 35 years..

anything over 10% is considered to dicey. a failure does not mean you are broke if you live another few years , it means the income stream would be reduced whether you needed the money or not..

so this is how age/time frame works out .

[Robyn55]

Originally Posted by mathjak107

you are not following .

typically retirement planning standardizes on planning to 95 .. if you are first retiring and 80 years old your planning time frame would be about 15 years.

to have your money last 15-20 years is very different than the planning needed to last 30-40 years.

15 YEARS IS A LOT LESS DEPENDENANT ON SEQUENCES,INFLATION AND YEARLY RETURNS .

the longer you have to plan for the more critical those 3 issues become.
you still need to watch what you spend for shorter time frames but the 3 parameters are less critical to your outcome.

My father is 95 now - in good health - and his current life expectancy is 7 years. OTOH - his late mother lived to 104 - so who the heck knows?

Also - if my father doesn't die in his sleep - he will probably need a SNF that costs here - where I live - about $80k/year. In your neck of the woods - you're probably talking about $125k/year or more.

Just FYI - my father's sister - at age 93 - is living in an independent senior living facility in Westchester County. It costs about 30% more up there for a much smaller unit than my father has here (same owner).

You're talking one-sized-fits-all math that some 30 year olds are spouting - and I'm talking what older people I know are actually living through. Big difference IMO. Robyn

[Cameron60]

Question If a 62 year old single came into $1 million dollars and had nothing else but that money and a small SS at age 66, how would you allocate beginning tomorrow?

[mathjak107]

Originally Posted by Robyn55

My father is 95 now - in good health - and his current life expectancy is 7 years. OTOH - his late mother lived to 104 - so who the heck knows?

Also - if my father doesn't die in his sleep - he will probably need a SNF that costs here - where I live - about $80k/year. In your neck of the woods - you're probably talking about $125k/year or more.

Just FYI - my father's sister - at age 93 - is living in an independent senior living facility in Westchester County. It costs about 30% more up there for a much smaller unit than my father has here (same owner).

You're talking one-sized-fits-all math that some 30 year olds are spouting - and I'm talking what older people I know are actually living through. Big difference IMO. Robyn

i would disagree again:

the math is the math just as the facts are the facts . it does not change. people change and tailor things to their life style but the math never changes.

ten can equal 6+4 or 7+3 or 5+5 but it will always be 10.

there are facts that do not change .

all your examples are doing is showing that people need a cushion for unexpected expenses . which just reinforces why you need to know what your doing when you plan how you will get that income from your nest egg.

if your rate of withdrawal is to high for your allocations you will not have that buffer for the awe craps. trying to pull 4% from bonds in rising inflation will be one of those bad moves.

remember you are the one claiming none of this is needed and you are telling everyone the seat of their pants are fine for planning.

[mathjak107]

Originally Posted by Cameron60

Question If a 62 year old single came into $1 million dollars and had nothing else but that money and a small SS at age 66, how would you allocate beginning tomorrow?

how much income do they need and what is their stomach for volatility to get that income?

how reliable do they want that income stream to be?

there is no one size fits all .

there are are many factors to that question?

realize that these allocations are only to weed out what didn't work so well over and over in the past and nothing is 100% going forward.

it is only to get you tio the gate with the highest odds, life takes it from there as spending becomes dynamic to what is happening around you.

the numbers may say 40-50 grand is okay but if your health isn't so good or markets suck you may just naturally draw less...

[Robyn55]

Originally Posted by mathjak107

i would disagree again:

the math is the math just as the facts are the facts . it does not change. people change and tailor things to their life style but the math never changes.

ten can equal 6+4 or 7+3 or 5+5 but it will always be 10.

there are facts that do not change .

all your examples are doing is showing that people need a cushion for unexpected expenses . which just reinforces why you need to know what your doing when you plan how you will get that income from your nest egg.

if your rate of withdrawal is to high for your allocations you will not have that buffer for the awe craps. trying to pull 4% from bonds in rising inflation will be one of those bad moves.

remember you are the one claiming none of this is needed and you are telling everyone the seat of their pants are fine for planning.

Ten can equal 6+4 or 7+3 - but the main issue is whether you're going to need 10 or 8 or 12 or 20 now or X years down the road. We have a large degree of - but not total - control over this issue. Another issue is whether we'll have 10 or 8 or 12 or 20 X years down the road. I think we have little and sometimes no control over that issue. We're pretty much dealing with the whims of the markets.

For example - I don't think getting pretty old and needing more money as one ages (as opposed to less) is an unforeseeable event. It's something that has to be a part of financial planning. Which is why - even if you're 70 - depending on one's financial circumstances - it may be necessary to avoid indulgences today to pay for necessities tomorrow. Nothing "seat of your pants" about that.

Also - in this regard - I think one of the most important decisions a retiree can make is where to live. Where we live not only has direct costs (ranging from the actual price of housing to property taxes) - but lots of indirect ones as well. For example - we pay a lot less for things like our auto insurance - our Medigap policies - etc. than we'd pay even in south Florida. None of these things (or others) standing alone was determinative when we decided to move here. They're just the results of our decision (which was based on many things).

And the whole process is - of course - dynamic. I didn't decide to do X 10 years ago - and "set it and forget it". You may think that's "seat of your pants" - but we all have to observe what's going on around us. And modify things that might have worked 5 years ago - but aren't working now - and won't work in the foreseeable future. That's why any set in stone withdrawal/spending formula for retirement just won't work for most people IMO - except perhaps in a computer simulation.

FWIW - I don't think it's safe for most people to withdraw 4%+ under any financial scenario. And if/when inflation returns in a bigger way than it exists today - we may see a return to market correlations many people don't remember. E.g., that stocks go down when interest rates go up (and vice versa). The old "fed three steps and a stumble" rule. This correlation worked for decades pre-2000. But it has fallen apart since then. I wouldn't rule out its return in the future.

Note that one reason I'm so leery about any retirement scenario based 100% or in large part on computer simulations is because I have a fair amount of experience reading about/researching/developing market trading systems based on computer simulations. Most tend not to work in the real world at all. Even those that seem to work for a while tend to fall apart when under stress. Think Long Term Capital Management. With all those PhDs - Nobel prize winners - tons of data and computing power. Just going "poof" one day and almost bringing down world markets with it. You may think that not following systems based on computer simulations is "going by the seat of your pants". I don't. And I don't care what a computer is telling me if it's out of synch with what I'm observing in the real world.

I think perhaps the real difference between us is you and many others are looking for simple rules that will make our retirements bullet-proof financially. And I don't think that any such rules exist or can be formulated. Robyn

[mathjak107]

these are not computer simulations , the safe withdrawal numbers are based on actual worst case real returns over every single 30 year period.

the events that caused them do not matter only the results which are if you get less than a 1% average real return per year over any 15 year time frame you will most likely have to take a pay cut from whatever your safe withdrawal rate is and you will have little cushion left for the unexpected expenses..

it is not always 4% . the number is dependent on age and number of years you want to plan out to and allocations.

sure ,you can pull 4 ,5,6% from any amount of savings regardless of allocations but what amount of risk of being one of the ones who failed in the past do you want to chance?

that is all retirement planning with these studies tells you. if you want the maximum cushion for all those unknown expenses you speak of I certainly would run with something that failed less in the past.

as I said trying to pull 4 or 5% from bonds only, has failed over and over through out so many time frames.

betting your retirement that you will be the case where it was okay may be foolish.

again these were never meant to be life long rules. they are planning tools to get you in a relatively safe ball park for your savings amount and the amount you want to draw.

it is up to you to season to taste.

[Robyn55]

Originally Posted by mathjak107

these are not computer simulations , the safe withdrawal numbers are based on actual worst case real returns over every single 30 year period.

the events that caused them do not matter only the results which are if you get less than a 1% average real return per year over any 15 year time frame you will most likely have to take a pay cut from whatever your safe withdrawal rate is and you will have little cushion left for the unexpected expenses..

it is not always 4% . the number is dependent on age and number of years you want to plan out to and allocations.

sure ,you can pull 4 ,5,6% from any amount of savings regardless of allocations but what amount of risk of being one of the ones who failed in the past do you want to chance?

that is all retirement planning with these studies tells you. if you want the maximum cushion for all those unknown expenses you speak of I certainly would run with something that failed less in the past.

as I said trying to pull 4 or 5% from bonds only, has failed over and over through out so many time frames.

betting your retirement that you will be the case where it was okay may be foolish.

again these were never meant to be life long rules. they are planning tools to get you in a relatively safe ball park for your savings amount and the amount you want to draw.

it is up to you to season to taste.

Of course they're simulations - because no one in the studies actually ever did it.

And you kind of got me interested. So I went and looked at what data these computer simulations are using. Apparently only the 10 year note (which has been around for a long time - but which was for the longest time something only available to institutional size investors). *And* the return on the SP500. Now the SP500 was only introduced in 1957 (its first predecessor - which didn't have 500 stocks - was introduced in 1923).** And it would have been impractical/impossible for individual investors to own the SP500 before 1976 - when Vanguard introduced the first SP500 mutual "index" fund (the SP500 ETF - SPY - was only introduced in 1993).

IOW - based on the available data - assuming you're comparing apples with apples - one can only make computations from 1957 (introduction of the SP500) to 1983 (the last full 30 year period ended in 2013). Which is only 27 time periods. Not exactly an overwhelming number of time periods.

FWIW - this is why I always hated "black box" trading systems. When someone (especially me!) is relying on data to do something - I want to know what the data is.

Also - Pfau is making certain assumptions about *future* market returns in his projections:

As a result [of current financial asset yields/prices], Pfau estimates a fifty-fifty portfolio of stocks and bonds is likely to deliver a long-run annual average return after inflation of just 2.2%, less than half thAs a result, Pfau estimates a fifty-fifty portfolio of stocks and bonds is likely to deliver a long-run annual average return after inflation of just 2.2%, less than half the historical rate.

And kind of assuming facts not in evidence:

Of course they're simulations - because no one actually ever did it.

And you kind of got me interested. So I went and looked at what data these computer simulations are using. Apparently only the 10 year note (which has been around for a long time - but which was for the longest time something only available to institutional size investors).* *And* the return on the SP500. Now the SP500 was only introduced in 1957 (its first predecessor - which didn't have 500 stocks - was introduced in 1923). And it would have been impractical/impossible for individual investors to own the SP500 before 1976 - when Vanguard introduced the first SP500 mutual "index" fund (the SP500 ETF - SPY - was only introduced in 1993).

IOW - based on the available data - one can only make computations from 1957 (introduction of the SP500) to 1983 (the last full 30 year period ended in 2013). Which is only 27 time periods. Not exactly an overwhelming number of time periods. Of course - if you trust the "blended" SP data - you can go back to 1928 - which gives you another 29 time periods. For a total of 56 annual time periods. I don't think that's enough data to base my retirement on.

FWIW - this is why I always hated "black box" trading systems. When someone (especially me!) is relying on data to do something - I want to know what the data is.

Also FWIW - Pfau is making certain assumptions about *future* market returns in his projections:

As a result [of current financial asset yields/prices], Pfau estimates a fifty-fifty portfolio of stocks and bonds is likely to deliver a long-run annual average return after inflation of just 2.2%, less than half the historical rate.

And assumptions about asset price correlations:

Even given the same long-run returns, a less volatile portfolio will tend to last longer. A portfolio that's not just large-cap stocks plus bonds, but is instead spread out to include small-company stocks, foreign stocks, and other asset classes, Pfau says, lets the safe drawdown from $1 million in savings rise to an initial $35,000 from $30,000.

Forget the 4% withdrawal rule - Feb. 26, 2014

Small cap stocks? The Russell 2000 first came out (as an index) in 1984. I buy/hold/trade the ETF (IWM) - which has only been around since 2000. We forget how new some of this stuff is. I sure don't expect the IWM to go up when the SP500 is going down. Indeed - in recent years - more and more asset classes that were traditionally non-correlated have become increasingly correlated:

RICH BERNSTEIN: 3 Things You Should Know About Financial Asset Correlation - Business Insider

The primary substantially non-correlated equities asset today is treasuries. OTOH - until recently - these assets tended to be correlated - a relationship that might reappear if interest rates were to rise a fair amount

Finally - a lot of these studies/this data - at least on the surface (perhaps it's all buried in footnotes) ignore a lot of important things. Like what do people do with interest/dividends? Spend or reinvest? If the latter - at what prices? What tax rates do people pay when they receive interest/dividends or sell assets (IRAs were first introduced in 1974). People have been known to harvest both capital gains and capital losses at times . What about commissions? Do these studies assume "frictionless" investing (minus taxes and fees)? If so - they're not realistic. Robyn

*Using only the 10 year note as the "bond proxy" is problematic for several reasons IMO. First - individual investors usually don't buy 10 year notes (and they're much much easier to buy today than in the past - and we're not talking ancient past either). Most non-institutional investors bought CDs - and usually shorter term ones. The yield was often better - but buying short term CDs leaves one unduly exposed to reinvestment risk. Individual investors with more money tended to buy munis - which have almost always offered better/much better after tax yields than the 10 year.

**This is the best history of the SP500 I could find:

History and Structure of the Standard & Poor's 500

The new SP500 data was tacked on to the data from an older different index so a body of historical data could be created. Whether this is good or bad - I don't know. All I know is you're dealing with data for 2 different indices.

[mathjak107]

to dateThe biggest issue one can raise is it is only 2 investments . to me that is a bigger concern than the fact they used a 10 year bond as there is no real history for what actual folks use..

Other investments were not tracked accurately or didn't exist so testing had to be
with what ever was readily available and trackable.

So for that matter few today only use two asset classes.

But , it still gives a very good insight into what worked and what didn't.

William Bengen’s work and the Trinity study used different bond indices. While Mr. Bengen’s original research combined the S&P 500 index with 5-year intermediate term government bond returns, the Trinity Study used long-term high-grade corporate bond returns instead.


PFAU'S prediction works just fine according to Michael KITCES as every time frame that failed produced less than 1% real return as an average for the first 15 years. . PFAU'S 2% is fine .

while at 4% the income stream will hold the amount left over in the bucket at the end will be a lot less if 2% returns are the norm.

if you are worried about the unexpected need for money popping up I would not want to draw 4% I would cut back to 3% to keep more in the bucket for the unexpected.

what you have to remember is these numbers that the safe withdrawal rates are based on are sooooooo conservative. they are not based on averages of any kind.. THEY ARE BASED ON THE WORST OUTCOMES EVER.


even moderately lousy would be an upside surprise to things.

the idea is the income stream can be reliable and consistent while bad markets and events play havoc with the amount in the bucket that will be left over. .

that bucket left is all over the place but the stream of income just remains unchanged.

to date that bucket has ended up with more in it at the end more than 90% of every time frame then you started with.

that is one big cushion for human nature and events to happen.